When all hypotheses are right: a multifocal account of dyslexia